A. Ibrahim, E., M. Youssef, A., Abdel-Hakim, M. (2006). AN ON-LINE OPTIMAL ARTIFICIAL NEURAL NETWORK-BASED CONTROLLER FOR SIMPLIFIED ORDER POWER SYSTEMS. JES. Journal of Engineering Sciences, 34(No 1), 89-106. doi: 10.21608/jesaun.2006.110083
El-Nobi A. Ibrahim; Ali. M. Youssef; M. Abdel-Hakim. "AN ON-LINE OPTIMAL ARTIFICIAL NEURAL NETWORK-BASED CONTROLLER FOR SIMPLIFIED ORDER POWER SYSTEMS". JES. Journal of Engineering Sciences, 34, No 1, 2006, 89-106. doi: 10.21608/jesaun.2006.110083
A. Ibrahim, E., M. Youssef, A., Abdel-Hakim, M. (2006). 'AN ON-LINE OPTIMAL ARTIFICIAL NEURAL NETWORK-BASED CONTROLLER FOR SIMPLIFIED ORDER POWER SYSTEMS', JES. Journal of Engineering Sciences, 34(No 1), pp. 89-106. doi: 10.21608/jesaun.2006.110083
A. Ibrahim, E., M. Youssef, A., Abdel-Hakim, M. AN ON-LINE OPTIMAL ARTIFICIAL NEURAL NETWORK-BASED CONTROLLER FOR SIMPLIFIED ORDER POWER SYSTEMS. JES. Journal of Engineering Sciences, 2006; 34(No 1): 89-106. doi: 10.21608/jesaun.2006.110083
AN ON-LINE OPTIMAL ARTIFICIAL NEURAL NETWORK-BASED CONTROLLER FOR SIMPLIFIED ORDER POWER SYSTEMS
Electrical Engineering Department, Faculty of Engineering, Assiut University, Assiut, Egypt
Abstract
The paper presents an on-line optimal artificial neural network (ANN)- based controller for simplified order power systems to improve the dynamic response under different operating conditions. The original 13th order power system is reduced to 5th order model. The basic feature of the proposed ANN controller is that it consists of two neural networks, one of them (ANN1) maps the optimal control process at different loading conditions and the other (ANN2) maps the feedback control to produce the required control action signal. The ANN1 is trained using input/output pairs of data which are collected from the optimal control of the reduced order model of power system at different loading conditions The ANN2 parameters are adapted on-line through the ANN1 according to loading conditions. The digital simulation results proved the high performance of the synchronous generator using the proposed ANN controller in terms of fast response and less undershot/overshot under different operating conditions. A comparison between the off-line fixed parameters optimal controller and the proposed ANN controller validates the effectiveness and reliability of the ANN controller.
View on SCiNiTO
Top